Environmental ImpactAssessment

Methodologies: Weighting factors, Utility functions,Quality indices and Multi-Attribute-Theories. Quantitativemethods. The Analytic Hierarchy Process (AHP).Evaluation and Comparison of Alternatives

Julius SolnesAutumn 2017

EIA methodology continued

PWeigthed checklists and environmental grades

P The Battelle Evaluation System

PMulti-Attribute-Utility theories< Utility functions< Environmental Quality Index< The AHP (?The Analytic Hierarchy Process” by Thomas

Saaty)

PA Case Study. The 685 MW Kárahjnúkar Power Plantand ALCOAs aluminium smelter in Reyðarfjörður

Weighted average of environmental impacts

P Sondheim (1978) uses weighted checklists (weighting panelsconstituted for each EIA)

P Yapijakis (1983) uses an adaption of Sondheim methods< Permanent national weighting panels with members from

academia, industry and government< In a trial project (river alteration, water supply), the national

panels of Greece and the former Yugoslavia (Macedonia)produced similar weights

< PROALT=wBCRqBCR+wCOqCO+wEIqEI+wSIqSI+wMTLqMTL� PROALT is the project alternative ranking or grade� BCR is the benefit/cost ratio of the alternative� CO is the total capital outlay for the alternative� EI is the quantified total environmental impact of alternative� SI is the social impact of alternative� MTL is the manageability and technology level of alternative� wI are the weights, decided by the weighting panels

Evaluation of alternativesCertain basic steps

P Determine an appropriate set of environmental factors to beconsidered (for example, wildlife habitat)

P Determine the environmental impact index for each factor< Define the units of measurement for each environmental factor (e.g.,

hectares preserved)< Collect data on the environmental factor (e.g., 10.000 hectares

preserved)< Decide on a common interval/index scale for each environmental

factor (e.g., 0 S 1, 1 S 5, or 0 S 10, sometimes 1 S 9 only)< Convert data for the environmental factor to an environmental factor

index (this is usually done by normalizing all values over a maximum orminimum value)

P Determine a weight (importance), wEF, for each environmentalfactor EF

P Decide on the method of aggregation across all factors (usuallyadditive)

Weighting and scaling techniques

Consider two environmental factors: wildlife habitat (hectares preserved)and employment increase (jobs created). EF’s are scaled to an index (0 isworst and 1 is best). Two different approaches: a) simple addition of factorindices, EF’s equally weighted; b) Wildlife habitat, weight 0.2, employmentincrease, weight 0.8.

Comparison result

Scaling: 10.000 ha = 1.0; 5000 jobs = 1.0

The Battelle Evaluation SystemThe Battelle Environmental Evaluation System (BEES) was developed by an inter-disciplinary research team at Battelle Columbus Laboratories, Ohio (Dee et al., 1972,1973). It is based on a hierarchical assessment of environmental quality indicators.

P Human Concern is speparated into four main categories.< Ecology (nature and natural habitats)< Physical/chemical< Aesthetics (visual impacts and more)< Human interest/social interest (f.inst. jobs and recreation possibilities)

P Each category is broken down into number of environmental components.

P For each component, an index/grade of environmental quality, normalizedto a scale ranging from 1 to 10, is developed, using value or utility functions

P Environmental indicator defined as difference in environmental qualitybetween "before" and "after" impact states.

P Each environmental component has a weighting factor (relativeimportance).

P Weights are fixed, and overall impact of project alternative is calculated bysumming the weighted impact indicators.

Value or Utility FunctionsThe Battelle Procedure for creating utility functions

P Step 1: Obtain information on the relationship between the parameter,EF, and quality of the environment.

P Step 2: Order the parameter scale (abscissa x) so that the lowest value iszero.

P Step 3: Divide the quality scale (ordinate y) into equal intervals between0 to 1, and determine the appropriate value of the parameter for eachinterval. Continue the process until a curve can be drawn.

P Step 4: Ask several different specialists to repeat Steps 1 to 3independently. Average/smooth the curves to obtain a group curve. (Forparameters based solely on value judgements, use a representativecross-section of the population.)

P Step 5: Show curves to all participants, and ask for a review if there arelarge variations. Modify the group curve as appropriate.

P Step 6: Repeat Steps 1-5 with a separate group of specialists, to test forreproducibility.

P Step 7: Repeat Steps 1-6 for all the selected parameters

Utility functions (Nytjaföll)Utility functions, enviromental ranking on a scale 0-10

Hypothetical utility function Ui(x) showingenvironmental ranking of percentage loss ofsalmonid populations in a river

0% 25% 50% 75% 100%

5

0

10

Percentage loss

Ui(x)

x

20 40 60 80 100%

Percent of carrying capacity based on animal units

8

6

4

3

0

10

Terrestrial carrying capacity for browsersand grazers

Battelle Evaluation System (Pros and Cons)

P Pros< Has high capability for identification and prediction of impacts, and

good replicability of results.< Provides high level of detail for assessment and documentation

purposes.< Basis for the development of environmental indicators, and associated

weights is fully documented.< Objectivity is high in terms of comparison between alternatives or

different project options.

P Cons< System is applicable only to projects for which it was designed;

development of additional indicators is demanding.< System has no mechanism for estimating or displaying interactions

between environmental components.< System does not link impacts to affected parties or to dominant issues.< The system has very high resource requirements (money, time,

manpower skills).

Environmental Quality Indices(Umhverfisgæðavísitölur)

Environmental indicesP Environmental Media Indices< Air quality (weighted average content of pollutants). Bad and good

ozone (O3), Mexico City example.< Water quality (weighted average content of pollutants)< Noise

P Ecological sensitivity and diversity� to evaluate the relative sensitivity to perturbations of the

ecosystems in a region

P Archeological resources� 19 parameters studied to evaluate the best waterway navigation

route through an archeologically sensitive region

P Visual quality

P Quality of life Index (QOL-vísitölur)

Multi-Attribute-Utility-Theories(Margfeldisáhrifa-nytjafræði)

Morgenstern and Neumann (1953), Keeney and Raiffa (1978)

P Multi-Attribute-Utility Theory

P Utility functions or value functions (nytjaföll) have to be defined foreach environmental factor (similar to the Battelle system)< The “Utility” (notagildi) of an environmental factor is measured on a scale 0-

10 (grading), where 10 is the “best” value (full utility), 0 is the “worst” value,e.g. total loss of certain environmental quality (zero utility)

P When the utility functions have been determined, it is possible tocombine the impacts through weighting (the weights orimportance factors ki). In this manner an Environmental QualityIndex, EQI („umhverfisvísitala“), is obtained:

Analytic Hierarchy Process (AHP-aðferðin)

Both the weights wi and the utility function values (scores) can beassessed, using the Analytic Hierarchy Process (AHP)

The Analytic Hierarchy Process was proposed by ThomasSaaty (a mathematician/economist in Chicago) in 1983, inorder to evaluate different investment alternatives

It is based on simple comparison matrices, where eachenvironmental factor is compared for all alternatives on aqualitative basis. The weights are obtained in the samemanner.

A more precise description can be found in severalOperation Analysis textbooks. See also Karahnjukar-AHP.pdfon the course website

Pairwise comparison matrix

The elements aij of the comparison matrix have to satisfy thefollowing conditions: aij=1/aji, aii=1, aij@ajk=aik. Consistency! CR<0,1for n>3 and <0.08 for n#3 (see Karahnjukar-AHP.pdf)

EF1 EF2 EF3 EF4 EF5 EF6 EF7

EF1

EF2

EF3

EF4

EF5

EF6

EF7

1

1

1

1

1

1

1

4

1/4

Comparison of environmental factors, weighting: EF i

EF5 is 4 times more important than EF4

Normalizing the comparison matrix

Step 1. For each column in the matrix, do the following: Find thesum of each column. Divide each entry in column i by the sum ofentries in that column. Repeat for all columns.

Finding the proper weights/importancefactor ki

Step 2. To find the approximate weight (importance) wi

for environmental factor EFi, calculate the average value ofthe entries in row j in Anorm, for instance the 5th row:

w5 is the best estimate of the proper weight, satisfyingthe conditions: aij=1/aji, aii=1, aij@ajk=aik

A Case historyEvaluation of three industrial development alternatives for the East Coast ofIceland

P Industrial and regional development of the underpopulated EastCoast of Iceland< a political and sociological dilemma

P Three possible Industrial Development Solutions< 340 thousand tons aluminium smelter plus massive 685 MW

hydropower development (Kárahnjúkavirkjun)< 6 million tons oil refinery without hydropower development< General government/private sector aided development with a

development fund

P Survey of general environmental consequences of all threealternatives

P Environmental Comparison of the three alternatives< Multi-attribute-utility-theory< Analytic Hierarchy Process Methodologies (AHP)< Environmental Quality Index (EQI)

Development of the underpopulatedEast Coast Region of Iceland

A political and social dilemma

Total population of Iceland:

330 thousand population units

Reykjavík (capital region):

200 thousand population units

East Coast population:

11 thousand population units

Reyðarfjörður population:

1500 population units

Iceland: 1 unit=1.0; China: 1 unit=4000

Aluminium Smelter with hydropowerdevelopment-ALT 1

Environmental Consequences

P Extensive land use with massive disruption of virgin mountainhighlands. Irreversible environmental consequences

P Negative social impact; environmental friendly image (-);ecotourism (-)

P Negative economic impact for traditional industries in the areabut positive impact for the area as a whole (??). In particular theplace where the smelter is located.

P Positive social and economic impact (+). Many jobs created withfiscal improvement for both local governments and persons.

P 30% increase of Icelandic CO2 emissions. Covered by the“Icelandic Clause” decided at COP 7 in Marrakesh 2001.

P Industrial pollution of hitherto pristine environment.

A six million tons oil refinery inReyðarfjörður-ALT 2

Environmental Consequences

P Oil spill accidents entail considerable environmental risk

P Negative social impact; environmental friendly image; ecotourism

P Negative economic impact for traditional industries in the area butpositive impact for the region as a whole(??). In particular the placewhere the refinery is located.

P Positive social and economic impact (+). Many jobs (high level) createdwith fiscal improvement for both local governments and persons.

P 30% increase of Icelandic CO2 emissions, which violates the Kyoto (Kjó-tó) protocol agreement (no-use of green energy, and not covered bythe “Icelandic Clause”)

P Limited industrial pollution

P Limited land use (about 50 ha coastal site)

General government/private sector aideddevelopment-ALT 3

Environmental Consequences

P Positive social impacts. Some possible job creation in newindustries (excotic products based on area’s natural resources)

P Conservation of the environmental friendly image of the region

P Ecotourism and safari/hunting expeditions; mountain lodges

P Traditional industries in the area can be expected to continuesuffering hard times

P Difficult but perhaps not impossible to establish informationtechnology firms, hitech and skills based companies

P Economic impacts weaker than for the industrial developmentsolutions, but more equally distributed throughout the region

P Airborne and industrial pollution continues at current low levels

Environmental Comparison of the ThreeDevelopment Alternatives

Environmental Factors Considered as Multi-Attribute-Utilities

P Land Use (industrial and hydropower sites,reservoirs and overhead lines): (LU)

P Economic Impacts: (EI)

P Social Impacts: (SI)

P Consequences for the natural environment: (CNE)

P Industrial Pollution: (IP)

P Greenhouse Gas Emissions: (GGE)

P Environmental Risks: (ER)

The AHP Comparison Matrix

Comparison of effects on scale 1-9; aij=1/aji, aii=1, aij@ajk=aik

The AHP Scores

SCORES are on a scale 0 (worst) to 10 (best)

LU: 2nd last row: ALT 3 (Gen. Developm.) is nine times better than ALT1 regarding land useEI: 2nd last row: ALT 3 (Gen. Developm.) has only 1/4 as much econonomic benefit as ALT 1

Environmental Quality Index for the threedevelopment alternatives

EQI calculated using the AHP methodology

ConclusionsP The Analytic Hierarchy Process offers a good methodology for

environmental comparison of alternatives

P As expected the “green” alternative has the highest EQI (6,16)

P The two industrial alternatives receive a very low score or EQI.2

P Faced with a choice between the two industrial alternatives, the oilrefinery seems only slightly favourable in terms of the EQI (2,19 vs2,00)

P The oil refinery has the advantage that the interior highlands are leftuntouched for future generations to enjoy

P The Government of Iceland should perhaps have studied the “green”alternative, before deciding to go for the smelter and the hugeKarahnjukar power plant. The profitability of the plant has beenseverely criticized for being unacceptable (energy price too low), andthis seems to be justified by recent disclosure of records by theowner, National Power Corporation (Landsvirkjun)

Multi-Attribute-Utility Theories General comments

Margfeldisáhrifa-nytjafræði; almennt um aðferðina

P Positive Points (kostir)< Well suited for engineers and natural scientists< Easy to apply probabilistic methodologies, e.g.

P[U(X)#u]=FU(u)< Possible to introduce “fancy” mathematical tools, e.g.

dynamic programming („aðgerðargreining“) or non-linearoptimization („ólínuleg beztun“)

< EXPERT CHOICE, an available commercial software

P Negative Points (Gallar)< The public looses insight into what is going on. Also likely

that a large part of the Assessment Team will not be happyor satisfied with such methods

Why did we (the Icelandic Government)choose the aluminium smelter

Although CO2 emission would exceed 500.000 tonsper annum, and thus technically speaking be inviolation of the Kyoto (kjó-tó) Protocol, Icelandwas allowed 10% increase of its emissions duringthe first period 1990-2012 (total CO2 emission1990 about 2.7 million tons), i.e. 270.000 tons.

The 500.000 tons p.a. will grossly exceed this limit.

Position of the Icelandic Government1995-2009

?The progress made in Iceland before 1990 to reduce the useof fossil fuels and carbon dioxide emission from local sourcesshould be acknowledged by the international community. Itshould therefore be recognized that latitude for furtherreduction of emissions from moveable sources in the countryis very limited. Moreover, the government considers anyemission reduction, which prevents the use of its “green” andenvironmental friendly energy resources, adverse to the aimsof the Kyoto Protocol. In this context it should be borne inmind that industrial development (meaning aluminiumsmelters) in Iceland will generate much less CO2 emissionsthan elsewhere, where fossil fuel will be the energy source”(very curious statement!)

Unharnessed Hydropower in the World

World Energy Council, Global Report: 2007 Survey of Energy Resources

Economically feasibleunharnessed hydropowerin some European countries:

Russia 852 Twh/yearNorway 187Sweden 85France 70Italy 65Sviss 41Iceland 40

Economically feasibleunharnessed hydropower in thedifferent continents:

Europe 2714 TWh/yearAfrica 1852North America 3012South America 3036Asia 5523Middle East 168Oceania 189

However, it was at that time claimed by the authoritiesthat only in Iceland can aluminium smelters be built using“green” energy!!!!!

The Marrakesh Meeting or COP 7 in 2001

Ministerial meeting of the FCCC countries in November 2001

P More than 80 ministers for the environment, topgovernment officials, and NGO’s attended

P The main result can be seen on the website of the FCCCSecretariate in Bonn: http://unfccc.int. Environmentalists agreed that the “Earth”, i.e. ourclimate, was the looser. In order to reach an agreement,most countries were accorded any provisions orconditions they had laid down, and all objectionsconsidered.

P The so-called Icelandic exemption clause was approved,subject to certain conditions.

The Icelandic ClauseImportance of “green” energy utilization acknowledged

P Covers new energy intensive industrial facilities in one location orenlargement of existing facilities after 1990

P If such industrial facility (new or enlargement) increases the totalannual emissions (reference year 1990) more than 5% within the firstProtocol phase 2001-2007, it may be exempted

P Conditions< The total emissions of a country is less than 0.05% of that of all industrial

nations 1990< Only renewable energy shall be used (hydropower, geothermal), which

leads to less global emissions per production unit< Best possible pollution control and best available technology to prevent

greenhouse gas emission shall be used

P Such emission, due to large industrial facilities in one country, Þ 1,6Mtons CO2/year

ENDIR

See you next Thursday. Last lecture: The Reykjavikdomestic airport. Relocation or let it stay!Invite friends who are interested in the relocationof Reykjavik Airport

Profitability of the 685 MW Kárahnjúkavirkjun

P An avid spokesman for hydropower development confessesthat the building of a large hydropower plant on thesepremises is more like a social project, that is, it does notnecessarily have economic benefit rather social benefit only

P Initial cost 1.2 billion dollars (685 MW), 2003 prices

< Annual costs� Amortization (30 year loan at 6,5%): 78 million USD� Operation cost (1% af stofnkostnaði): 12 million USD

� Total 90 million USD

< Annual max. income (Landsvirkjun refuses to divulge the unit priceof electricity (possibly as low as 18 Mils/kWh)). Annual productioncapacity 4450 Gwst. Annual sale (4200 Gwst. á 2,0 cent/KWst)could amount to 84 million USD, i.e. negative profitability.

P The aluminium smelter has never paid any taxes in Iceland!